MLLGApr 16, 2019

Risk Bounds for Learning Multiple Components with Permutation-Invariant Losses

arXiv:1904.07594v24 citations
Originality Incremental advance
AI Analysis

This work provides improved theoretical guarantees for machine learning problems like switching regression and clustering, though it appears incremental as it builds on existing complexity decompositions.

The paper tackles the problem of deriving efficient error bounds for learning multiple components with sparsity-inducing regularization, showing that the approach yields risk bounds with mild dependencies on the number of components and removes this dependence for nonconvex regularization schemes.

This paper proposes a simple approach to derive efficient error bounds for learning multiple components with sparsity-inducing regularization. We show that for such regularization schemes, known decompositions of the Rademacher complexity over the components can be used in a more efficient manner to result in tighter bounds without too much effort. We give examples of application to switching regression and center-based clustering/vector quantization. Then, the complete workflow is illustrated on the problem of subspace clustering, for which decomposition results were not previously available. For all these problems, the proposed approach yields risk bounds with mild dependencies on the number of components and completely removes this dependence for nonconvex regularization schemes that could not be handled by previous methods.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes